From Seeing to Doing: Bridging Reasoning and Decision for Robotic Manipulation

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Yuan, Yifu, Cui, Haiqin, Chen, Yibin, Dong, Zibin, Ni, Fei, Kou, Longxin, Liu, Jinyi, Li, Pengyi, Zheng, Yan, Hao, Jianye
Format: Preprint
Published: 2025
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866917383283671040
author Yuan, Yifu
Cui, Haiqin
Chen, Yibin
Dong, Zibin
Ni, Fei
Kou, Longxin
Liu, Jinyi
Li, Pengyi
Zheng, Yan
Hao, Jianye
author_facet Yuan, Yifu
Cui, Haiqin
Chen, Yibin
Dong, Zibin
Ni, Fei
Kou, Longxin
Liu, Jinyi
Li, Pengyi
Zheng, Yan
Hao, Jianye
contents Achieving generalization in robotic manipulation remains a critical challenge, particularly for unseen scenarios and novel tasks. Current Vision-Language-Action (VLA) models, while building on top of general Vision-Language Models (VLMs), still fall short of achieving robust zero-shot performance due to the scarcity and heterogeneity prevalent in embodied datasets. To address these limitations, we propose FSD (From Seeing to Doing), a novel vision-language model that generates intermediate representations through spatial relationship reasoning, providing fine-grained guidance for robotic manipulation. Our approach combines a hierarchical data pipeline for training with a self-consistency mechanism that aligns spatial coordinates with visual signals. Through extensive experiments, we comprehensively validated FSD's capabilities in both "seeing" and "doing," achieving outstanding performance across 8 benchmarks for general spatial reasoning and embodied reference abilities, as well as on our proposed more challenging benchmark VABench. We also verified zero-shot capabilities in robot manipulation, demonstrating significant performance improvements over baseline methods in both SimplerEnv and real robot settings. Experimental results show that FSD achieves 40.6% success rate in SimplerEnv and 72% success rate across 8 real-world tasks, outperforming the strongest baseline by 30%.
format Preprint
id arxiv_https___arxiv_org_abs_2505_08548
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle From Seeing to Doing: Bridging Reasoning and Decision for Robotic Manipulation
Yuan, Yifu
Cui, Haiqin
Chen, Yibin
Dong, Zibin
Ni, Fei
Kou, Longxin
Liu, Jinyi
Li, Pengyi
Zheng, Yan
Hao, Jianye
Robotics
Artificial Intelligence
Machine Learning
Achieving generalization in robotic manipulation remains a critical challenge, particularly for unseen scenarios and novel tasks. Current Vision-Language-Action (VLA) models, while building on top of general Vision-Language Models (VLMs), still fall short of achieving robust zero-shot performance due to the scarcity and heterogeneity prevalent in embodied datasets. To address these limitations, we propose FSD (From Seeing to Doing), a novel vision-language model that generates intermediate representations through spatial relationship reasoning, providing fine-grained guidance for robotic manipulation. Our approach combines a hierarchical data pipeline for training with a self-consistency mechanism that aligns spatial coordinates with visual signals. Through extensive experiments, we comprehensively validated FSD's capabilities in both "seeing" and "doing," achieving outstanding performance across 8 benchmarks for general spatial reasoning and embodied reference abilities, as well as on our proposed more challenging benchmark VABench. We also verified zero-shot capabilities in robot manipulation, demonstrating significant performance improvements over baseline methods in both SimplerEnv and real robot settings. Experimental results show that FSD achieves 40.6% success rate in SimplerEnv and 72% success rate across 8 real-world tasks, outperforming the strongest baseline by 30%.
title From Seeing to Doing: Bridging Reasoning and Decision for Robotic Manipulation
topic Robotics
Artificial Intelligence
Machine Learning
url https://arxiv.org/abs/2505.08548